CN111061948A - User label recommendation method and device, computer equipment and storage medium - Google Patents
User label recommendation method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN111061948A CN111061948A CN201911167962.5A CN201911167962A CN111061948A CN 111061948 A CN111061948 A CN 111061948A CN 201911167962 A CN201911167962 A CN 201911167962A CN 111061948 A CN111061948 A CN 111061948A
- Authority
- CN
- China
- Prior art keywords
- label
- target
- tag
- value
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000000875 corresponding effect Effects 0.000 claims abstract description 242
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 46
- 230000002596 correlated effect Effects 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 10
- 238000012163 sequencing technique Methods 0.000 claims abstract description 9
- 238000004364 calculation method Methods 0.000 claims description 31
- 238000004590 computer program Methods 0.000 claims description 14
- 230000006399 behavior Effects 0.000 claims description 12
- 230000002452 interceptive effect Effects 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 4
- 230000003993 interaction Effects 0.000 description 11
- 238000004891 communication Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 4
- 230000008451 emotion Effects 0.000 description 4
- 238000002372 labelling Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 241001672694 Citrus reticulata Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a user tag recommendation method and device, computer equipment and a storage medium, and belongs to the technical field of data processing. The method comprises the following steps: determining a label combination related to a business target from a label set of a target user; acquiring a weight value of each label in the label combination corresponding to a business target according to a first algorithm; acquiring a weight value of each label in the label combination corresponding to the target user according to a second algorithm; acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm; calculating the final weight value of each label based on the weight value of each label corresponding to the target service, the weight value of each label corresponding to the target user and the time attenuation coefficient corresponding to each label; and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner. The invention can avoid wasting server resources when the user label is recommended, and improve the effectiveness of label recommendation.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a user tag recommendation method and device, computer equipment and a storage medium.
Background
With the development of internet finance, more and more individuals or enterprises seek financial services such as loan and loan through the internet, and meanwhile, the number of overdue loan cases is increasing.
In order to facilitate the collection of the overdue loan case by the collector, the prior art generally obtains various information of the overdue loan applicant through a server, and recommends the obtained various information to the corresponding electric collector to assist the collector in collecting the loan. However, the existing server has limited resources, and if the server needs to repeatedly push too much information which is not suitable for the overdue loan applicant to the acquirer for each overdue loan applicant, the resource of the server is wasted, and the acquirer cannot obtain useful information in time due to the fact that the information recommended to the acquirer is too much and complicated, so that the acquisition prompting effect is affected.
Disclosure of Invention
In order to solve the technical problems mentioned in the background art, the invention provides a user tag recommendation method, a user tag recommendation device, a computer device and a storage medium, which can not only avoid wasting server resources when a user tag is recommended, but also improve the effectiveness of tag recommendation.
The embodiment of the invention provides the following specific technical scheme:
in a first aspect, a user tag recommendation method is provided, where the method includes:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to the service target according to a first algorithm;
obtaining a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating a final weight value of each label based on a weight value of each label corresponding to the target service, a weight value of each label corresponding to the target user and a time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
Further, when the collection hastening case associated with the target user is successfully collected as the business target, the determining a tag combination related to the business target from the tag set of the target user includes:
acquiring a plurality of basic labels, a plurality of behavior labels and a plurality of interactive labels of the target user from the label set to form a label combination;
when the call number associated with the target user is taken as the effective number as the service target, the determining of the label combination related to the service target from the label set of the target user comprises the following steps:
and acquiring a plurality of interactive labels of the target user from the label set to form the label combination.
Further, the obtaining, according to the first algorithm, a weight value of each label in the label combination corresponding to the service objective includes:
determining a plurality of groups corresponding to the labels respectively, and performing WOE calculation and IV calculation on the groups of the labels respectively to obtain IV values of the labels;
training a preset model by taking each label as an input variable and taking the service target as an output variable, and calculating to obtain a numerical value of the feature importance of each label;
and for each label, carrying out weighted summation on the IV value of the label and the numerical value of the characteristic importance of the label to obtain a weight value of the label corresponding to the business target.
Further, the obtaining, according to a second algorithm, a weight value of each tag in the tag combination corresponding to the target user includes:
for each of the tags, performing the following operations:
calculating the TF value of the label according to the number of times of the label and the total number of times of all the labels in the label combination;
calculating the IDF value of the label according to the total number of all users and the number of the users containing the label;
and calculating the weight value of the label corresponding to the target user according to the TF value of the label and the IDF value of the label.
Further, the obtaining a time attenuation coefficient corresponding to each tag in the tag combination according to a third algorithm includes:
sorting the weighted values of the labels in the label combination corresponding to the business targets from large to small, and setting corresponding first initial values for the labels respectively according to the sorting result;
determining the current time period of the service corresponding to the service target;
performing attenuation calculation on a first initial value corresponding to each label according to a preset attenuation value corresponding to the current time period of the service to obtain an initial coefficient corresponding to each label;
acquiring initial attenuation factor values corresponding to time periods of the tags, wherein the time period of any one tag is determined based on the number of days of time for marking the tag for the first time by the target user;
respectively setting corresponding second initial values for the labels according to the sorting result;
taking the second initial value corresponding to each label as an attenuation value, and performing attenuation calculation on the initial attenuation factor value corresponding to the time period in which each label is currently located to obtain the time attenuation factor value corresponding to the time period in which each label is currently located;
and calculating to obtain the time attenuation coefficient corresponding to each label according to the initial coefficient corresponding to each label and the time attenuation factor value corresponding to the time period in which each label is currently positioned.
Further, when the collection urging case associated with the target user is used as the service, the determining of the current time period of the service corresponding to the service target includes:
determining the current time period of the business according to the overdue days of the collection urging case;
when the call number associated with the target user is used as the service, the determining the current time period of the service corresponding to the service target includes:
and determining the current time period of the service according to the time days of the first connection of the call number.
Further, the method further comprises:
for each label, comparing the final weight value of the label with a preset threshold value;
and if the final weight of the label exceeds the preset threshold value, preferentially taking the label as the target label, and recommending the target label to a corresponding acquirer in association with the business target.
In a second aspect, there is provided a user tag recommendation apparatus, the apparatus comprising:
the label determining module is used for determining a label combination related to the business target from the label set of the target user;
the first obtaining module is used for obtaining the weight value of each label in the label combination corresponding to the service target according to a first algorithm;
the second obtaining module is used for obtaining the weight value of each label in the label combination corresponding to the target user according to a second algorithm;
the third obtaining module is used for obtaining the time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
a weight calculation module, configured to calculate a final weight value of each tag based on a weight value of each tag corresponding to the target service, a weight value of each tag corresponding to the target user, and a time attenuation coefficient corresponding to each tag;
and the label recommending module is used for sorting the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
Further, when the collection case associated with the target user is successfully collected as the business target, the tag determination module is specifically configured to:
acquiring a plurality of basic labels, a plurality of behavior labels and a plurality of interactive labels of the target user from the label set to form a label combination;
when the call number associated with the target user is taken as the valid number as the service target, the tag determination module is specifically configured to:
and acquiring a plurality of interactive labels of the target user from the label set to form the label combination.
Further, the first obtaining module is specifically configured to:
determining a plurality of groups corresponding to the labels respectively, and performing WOE calculation and IV calculation on the groups of the labels respectively to obtain IV values of the labels;
training a preset model by taking each label as an input variable and taking the service target as an output variable, and calculating to obtain a numerical value of the feature importance of each label;
and for each label, carrying out weighted summation on the IV value of the label and the numerical value of the characteristic importance of the label to obtain a weight value of the label corresponding to the business target.
Further, the second obtaining module is specifically configured to:
for each of the tags, performing the following operations:
calculating the TF value of the label according to the number of times of the label and the total number of times of all the labels in the label combination;
calculating the IDF value of the label according to the total number of all users and the number of the users containing the label;
and calculating the weight value of the label corresponding to the target user according to the TF value of the label and the IDF value of the label.
Further, the third obtaining module is specifically configured to:
sorting the weighted values of the labels in the label combination corresponding to the business targets from large to small, and setting corresponding first initial values for the labels respectively according to the sorting result;
determining the current time period of the service corresponding to the service target;
performing attenuation calculation on a first initial value corresponding to each label according to a preset attenuation value corresponding to the current time period of the service to obtain an initial coefficient corresponding to each label;
acquiring initial attenuation factor values corresponding to time periods of the tags, wherein the time period of any one tag is determined based on the number of days of time for marking the tag for the first time by the target user;
respectively setting corresponding second initial values for the labels according to the sorting result;
taking the second initial value corresponding to each label as an attenuation value, and performing attenuation calculation on the initial attenuation factor value corresponding to the time period in which each label is currently located to obtain the time attenuation factor value corresponding to the time period in which each label is currently located;
and calculating to obtain the time attenuation coefficient corresponding to each label according to the initial coefficient corresponding to each label and the time attenuation factor value corresponding to the time period in which each label is currently positioned.
Further, when the collection urging case associated with the target user is taken as the service, the third obtaining module is specifically configured to:
determining the current time period of the business according to the overdue days of the collection urging case;
when the call number associated with the target user is used as the service, the third obtaining module is specifically configured to:
and determining the current time period of the service according to the time days of the first connection of the call number.
Further, the tag recommendation module is further configured to:
for each label, comparing the final weight value of the label with a preset threshold value;
and if the final weight of the label exceeds the preset threshold value, preferentially taking the label as the target label, and recommending the target label to a corresponding acquirer in association with the business target.
In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to the service target according to a first algorithm;
obtaining a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating a final weight value of each label based on a weight value of each label corresponding to the target service, a weight value of each label corresponding to the target user and a time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
In a fourth aspect, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to the service target according to a first algorithm;
obtaining a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating a final weight value of each label based on a weight value of each label corresponding to the target service, a weight value of each label corresponding to the target user and a time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
The embodiment of the invention provides a user tag recommendation method, a device, a computer device and a storage medium, wherein each tag related to a business target is determined from tags of a target user in a centralized manner, a final weight value of each tag is obtained by calculation based on a weight value of each tag corresponding to the business target, a weight value of each tag corresponding to a user and a time attenuation coefficient corresponding to the tag, and effective tags are screened out by sequencing each tag according to the final weight value, so that a server can recommend a user tag with a higher final weight value to a receiver, the tag recommendation is more targeted and useful, excessive information which is inapplicable to an overdue loan applicant and is repeatedly pushed to the receiver by the server in the prior art is avoided, the waste of server resources is avoided, and the tag recommendation effectiveness is improved, meanwhile, more useful user tags are recommended to the collector, the loan collection efficiency can be improved, and the collection effect is further improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is an application environment diagram of a user tag recommendation method according to an embodiment of the present invention;
fig. 2 is a flowchart of a user tag recommendation method according to an embodiment of the present invention;
FIG. 3 is a detailed flowchart of step S22 of the method shown in FIG. 2;
FIG. 4 is a detailed flowchart of step S23 of the method shown in FIG. 2;
FIG. 5 is a detailed flowchart of step S24 of the method shown in FIG. 2;
fig. 6 is a structural diagram of a user tag recommendation apparatus according to an embodiment of the present invention;
fig. 7 is an internal structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It is to be understood that, unless the context clearly requires otherwise, throughout the description and the claims, the words "comprise", "comprising", and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is, what is meant is "including, but not limited to".
Furthermore, in the description of the present invention, it is to be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. In addition, in the description of the present invention, "a plurality" means two or more unless otherwise specified.
The user tag recommendation method provided by the invention can be applied to the application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The server 104 first determines a tag combination related to a service target from a tag set of a target user, then obtains a weight value of each tag in the tag combination corresponding to the service target according to a first algorithm, obtains a weight value of each tag in the tag combination corresponding to the user according to a second algorithm, obtains a time attenuation coefficient of each tag in the tag combination according to a third algorithm, calculates a final weight value of each tag based on the weight value of each tag corresponding to the target service, the weight value of each tag corresponding to the user, and the time attenuation coefficient corresponding to each tag, sorts each tag according to the respective final weight value, screens out an effective tag, and recommends the effective tag to the corresponding terminal 102 in association with the service target. The terminal 102 is a terminal of an acquirer, and may be but is not limited to various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices, and the server 104 may be implemented by an independent server or a server cluster formed by a plurality of servers. In addition, the method provided by the embodiment of the invention can be applied to a loan collection prompting scene, and effective tags are intensively screened out from tags of overdue loan applicants through the server and recommended to a collector, so as to assist the collection prompting work of collection prompting.
Fig. 2 is a flowchart of a user tag recommendation method according to an embodiment of the present invention, which is described by taking the method as an example of being applied to the server in fig. 1, and as shown in fig. 2, the method may include:
in step S21, a tag combination related to the business objective is determined from the tag set of the target user.
Specifically, after the server selects a certain user as a target user, the tag set corresponding to the target user may be searched according to the user identifier corresponding to the target user. The tag set of the user includes various tags of the user, and the various tags are all used for representing attributes of the user.
The method comprises the steps of labeling user basic information, user behavior information and user voice interaction information in advance to obtain user basic labels, user behavior labels and user interaction labels to form a label set of a user, wherein the user voice interaction information refers to telephone communication information of a receiver and an overdue user, and the labeling processing comprises but is not limited to statistical analysis, unsupervised method classification, prediction models and the like.
Specifically, the tagging process is performed on the user basic information, and the user basic tag can be acquired, for example: age, academic calendar, occupation, gender, region, marital, etc., the user behavior information (e.g., historical borrowing behavior information, historical repayment behavior information, online operation behavior information) is tagged, and a user behavior tag may be obtained, for example: whether a header mark is left over, a borrowing habit, a repayment habit, an online login habit and the like, processing the voice interaction information of the user to acquire a user interaction tag, for example: the communication emotion fluctuation degree, the communication emotion type, whether the number is effective or not, the strength degree, whether the communication is smooth or not and the like.
More specifically, for the user voice interaction information, the user voice interaction information can be converted into text information, important label information such as a new user number and the like is obtained from the text information through relevant rules, the validity of the call is judged according to a preset model for the text information, namely the possibility of payment is promoted, and the type of the call result is judged according to the rules for the text information; in addition, the communication emotion types such as anger, fear and neutrality can be identified from the voice interaction information of the user through a pre-trained model; in addition, the method can also acquire the fluctuation degree of the conversation emotion from the voice interaction information of the user, count the ratio of silence and the number of times of call robbery, judge the strength degree of the overdue user, and identify whether the voice interaction information of the user is dialect or mandarin so as to judge whether the communication is smooth; in addition, the speed, tone and volume information of the user can be identified from the voice interaction information.
The collection urging case related to the user can be used as a business, and correspondingly, the collection urging case related to the user is successfully collected as a business target, wherein the collection urging case refers to a case with overdue loan allocated to a collector, and the collector can be an employee or an intelligent collection urging robot. The term "urging" is understood to mean urging the case to return a successful fund. In addition, the call number associated with the user may also be used as a service, and accordingly, the call number associated with the user may also be used as an effective number as a service target.
Generally, the same label will often exhibit different effects for different business objectives, such as a learned calendar, the higher the learned calendar the easier it is to refund a case, but the higher the learned calendar the less significant the relationship between the validity of the number. Therefore, when the business target is the successful urging of the user-associated case for collection, the academic record is the label related to the business target; when the service objective is that the call number associated with the user is a valid number, the learned calendar is not a tag associated with the service objective.
The embodiment of the present invention does not specifically limit the specific determination process.
Step S22, obtaining a weight value corresponding to the service target for each label in the label combination according to the first algorithm.
Specifically, a weight value of each tag in the tag combination corresponding to the business target may be obtained by using a scoring card model. For example, when the call is successfully issued by the call-in case as a business target, the Information Value (IV) is used to measure the weight Value of each tag in the tag combination of the overdue user corresponding to the business target by analyzing the call-in records of a large number of overdue users, wherein the weight Value of the tag corresponding to the business target is used to represent the influence degree of the tag on the business target.
Step S23, obtaining a weight value corresponding to the target user for each tag in the tag combination according to a second algorithm.
Specifically, the TF-IDF algorithm or lda (late Dirichlet allocation) algorithm may be utilized to obtain the weight value corresponding to the target user for each tag in the tag combination. The weight value of the label corresponding to the target user can be used for representing the importance degree of the label to the target user.
Step S24, obtaining the time attenuation coefficient corresponding to each label in the label combination according to the third algorithm.
In this embodiment, considering that the tag is influenced by the time factor, the influence degree of the tag of the target user on the successful urging of the case of the target user (or the validity of the call number associated with the target user) is long and weak with time, and therefore, in order to reflect the influence factor, the time attenuation coefficient corresponding to each tag in the tag combination may be obtained according to the third algorithm.
Step S25, calculating a final weight value of each label based on the weight value of each label corresponding to the target service, the weight value of each label corresponding to the target user, and the time attenuation coefficient corresponding to each label.
Specifically, for each tag, a weight value corresponding to the target service of the tag, a weight value corresponding to the target user of the tag, and a time attenuation coefficient corresponding to the tag may be multiplied to obtain a final weight value of the tag.
The present embodiment does not specifically limit the specific calculation process.
And step S26, sorting the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
Specifically, the labels are sorted from large to small according to the respective final weight values to obtain a label sorted list, top N effective labels (where N is a positive integer) are sorted from the label sorted list, and the sorted effective labels are recommended to corresponding acquirers in association with the business target, where N may be set according to actual needs, for example, N is 30.
Further, before step S26, the method may further include:
and comparing the final weight value of each label with a preset threshold value, and if the final weight value of each label exceeds the preset threshold value, preferentially taking the label as a target label and recommending the target label to a corresponding acquirer in association with a business target.
The preset threshold value can be set according to actual needs, and after the target tags are screened out, the target tags and the business targets are recommended to corresponding collectors in an associated manner, and the collectors distribute the collecting cases associated with the target users in advance.
In this embodiment, usually, the tag whose final weight value exceeds the preset threshold value brings a substantial increase of risk, or a substantial increase of the probability of reimbursement, so that the tag whose final weight value exceeds the preset threshold value is preferentially associated with the service target and recommended to the corresponding acquirer, so that the acquirer can timely obtain the tag having a large influence on the service scenario, where the service scenario may be but is not limited to: the method has the advantages of urging to collect the fraud risk of cases, having extremely high urging probability of cases, the complaint risk of call numbers and the invalid communication of the numbers.
The embodiment of the invention provides a user label recommendation method, which comprises the steps of determining all labels related to a business target from a label set of a target user, calculating a final weight value of each label based on a weight value of each label corresponding to the business target, a weight value of each label corresponding to a user and a time attenuation coefficient corresponding to the label, and sorting the labels according to the final weight values to screen out effective labels, so that a server can recommend a user label with a higher final weight value to a receiver, the label recommendation is more targeted and useful, the situation that the server repeatedly pushes excessive information which is inappropriate for an overdue loan applicant to the receiver in the prior art is avoided, the waste of server resources is avoided, the effectiveness of label recommendation is improved, and meanwhile, more useful user labels are recommended to the receiver, the loan collection efficiency can be improved, and the collection effect is improved.
In a preferred example, when the hasty case associated with the target user is successfully induced to serve as the business target, the specific implementation process of the step S21 may include:
and acquiring a plurality of basic labels, a plurality of behavior labels and a plurality of interactive labels of the target user from the label set to form a label combination.
When the call number associated with the target user is taken as the valid number as the service target, the specific implementation process of step S21 may include:
and acquiring a plurality of interactive labels of the target user from the label set to form a label combination.
In this embodiment, a tag combination related to a business target is determined from a tag set of a target user, so that a subsequent server evaluates the influence degree of each tag on the business target.
In a preferred example, as shown in fig. 3, the specific implementation process of the step S22 may include:
and S31, determining a plurality of groups corresponding to each label, and performing WOE calculation and IV calculation on the plurality of groups of each label to obtain the IV value of each label.
Specifically, for each label, determining respective WOE values of a plurality of packets corresponding to the label, calculating respective IV values of the plurality of packets corresponding to the label according to the respective WOE values of the plurality of packets corresponding to the label, and performing cumulative summation on the respective IV values of the plurality of packets corresponding to the label to obtain the IV value of the label.
The IV value of the label is larger, the influence degree of the label on the business target is higher, and the corresponding label weight is also higher.
The following describes a specific implementation process of the IV value of each tag by taking the successful urging of the user-associated urging case as a business objective.
Determining n groups (n is an integer greater than 0) corresponding to each label through chi-square binning and manual adjustment, wherein for the ith group corresponding to any label, the WOE value is calculated according to the following formula:
wherein, pyiThe ratio of cases to be catalyzed in the group is shown in all cases; pn-N complexiThe ratio of cases which are not catalyzed in the group to all cases which are not catalyzed is shown; # yiMeans the number of cases in the group; # ytThe number of cases to be induced in all cases to be induced is shown; # niThe number of cases not yet induced in the group; # ntThe number of the cases which are not induced in all the cases which are induced is shown, i is a positive integer less than n and represents the ith group in n groups.
Correspondingly, for the ith group of any tag, the IV value is calculated as follows:
the IV value of the tag can be calculated according to the following calculation formula:
the above-mentioned occurrence case refers to a case in which the target variable in the score card model takes a value of "yes" or "1", and the above-mentioned non-occurrence case refers to a case in which the target variable in the score card model takes a value of "no" or "0".
And S32, training the preset model by taking each label as an input variable and taking the business target as an output variable, and calculating to obtain a numerical value of the feature importance of each label.
When the preset model is trained, the label variables are used as the input of the preset model, the label variables comprise non-numerical variables and numerical variables, the non-numerical variables comprise sex and occupation, and the non-numerical variables can be used as the input of the preset model after feature coding. The preset model may be a LightGBM model, a GBDT model, or a random forest model, which is not specifically limited in this embodiment of the present invention.
In this embodiment, feature importance (feature importance) can be used to evaluate the degree of influence of multiple tags on a business target.
And S33, for each label, carrying out weighted summation on the IV value of the label and the numerical value of the characteristic importance of the label to obtain a weight value of the label corresponding to the business target.
Specifically, for each label, the IV value of the label and the value of the feature importance of the label are respectively normalized to obtain a normalized IV value of the label and a normalized value of the feature importance of the label; aiming at any label T, obtaining the weight value W of each label corresponding to the business target according to the following calculation formulaT：WTIV value weight after normalization + numerical value of feature importance after normalization-feature importance weight; the IV value weight of the tag T and the feature importance weight of the tag T may be preset according to actual needs, for example, the IV value weight and the feature importance weight are respectively set to 0.6 and 0.4, which is not limited in this embodiment of the present invention.
In addition, the IV value of each tag and the value of the feature importance of each tag may be sorted to obtain the IV value of each sorted tag and the value of the feature importance of each sorted tag, and for any tag T, the weight value W of each tag corresponding to the business object may be obtained according to the following calculation formulaT：WTIV value weight + numerical value of feature importance after sorting.
In this embodiment, for each tag, a weight value of the tag corresponding to the business target is obtained by calculating according to the IV value of the tag and the numerical value of the feature importance of the tag, so that the weight value of the tag corresponding to the business target can more accurately reflect the degree of influence of the tag on the business target.
In a preferred example, as shown in fig. 4, the specific implementation process of the step S23 may include:
s41, for each tag, calculating the TF value of the tag according to the number of occurrences of the tag and the total number of occurrences of all tags in the tag combination.
Specifically, the ratio between the number of times the label appears and the total number of times all labels appear in the label combination is calculated, and the calculated ratio is used as the TF value of each label.
The number of times of label occurrence is equal to the number of times of label marking for the target user, and the total number of times of label occurrence in the label combination is equal to the sum of the number of times of label marking for the target user.
S42, calculating the IDF value of the label according to the total number of all users and the number of users including the label.
Specifically, for each tag, the number of all overdue users and the number of users including the tag may be counted, a ratio between the number of all overdue users and the number of users including the tag is calculated, and the calculated ratio is used as the IDF value of each tag.
And S43, calculating the weight value of the label corresponding to the target user according to the TF value of the label and the IDF value of the label.
Specifically, the TF value of each tag and the IDF value of each tag are multiplied correspondingly to obtain the TF-IDF value of each tag, and the TF-IDF value of each tag is used as the weight value of each tag corresponding to the user.
In this embodiment, the weight value of each label corresponding to the target user is calculated by using the TF-IDF algorithm, so that the target user can be well distinguished from other users.
In a preferred example, as shown in fig. 5, the specific implementation process of step S24 may include:
s51, sorting the weighted values of the labels in the label combination corresponding to the business targets from big to small, and setting corresponding first initial values for the labels respectively according to the sorting result.
Specifically, each tag in the tag combination may be divided into a plurality of levels according to a preset ratio according to the sorting result, and different first initial values may be set for tags in different levels. For example, the weight values of the labels corresponding to the business targets are sorted from large to small, the labels are divided into three levels according to a ratio of 3:4:3, different first initial values are set for the labels in each level, and for example, the first initial values set for the labels in the three levels are respectively 3,2 and 1. It will be appreciated that the earlier the level, the greater the initial value to which the tag in the level corresponds.
And S52, determining the current time period of the service corresponding to the service target.
Specifically, when a collection urging case associated with the target user is used as a service, the current time period of the service can be determined according to the number of overdue days of the collection urging case.
Specifically, the time period of the overdue days of the urge to collect cases is determined as the current time period of the business, wherein different attenuation values are preset in different time periods. For example, the time period is divided into: the attenuation values of the time-out period of 0 to 15 days, the time-out period of 16 to 30 days, the time-out period of 31 to 60 days and the time-out period of 61 days are respectively set to be 1, 0.9, 0.8 and 0.7.
In addition, when the call number associated with the target user is used as the service, the current time period of the service can be determined according to the number of days until the call number is connected for the first time. The time period of the current time days when the call number is first connected can be determined as the current time period of the service, which is not described herein again.
And S53, performing attenuation calculation on the first initial value corresponding to each label according to the preset attenuation value corresponding to the current time period of the service to obtain the initial coefficient corresponding to each label.
Specifically, a preset attenuation value corresponding to a current time period of the service is multiplied by a first initial value corresponding to each tag, so as to obtain an initial coefficient corresponding to each tag.
And S54, acquiring initial attenuation factor values corresponding to the current time periods of the labels.
Wherein the current time period of any label is determined based on the number of days until the label is first marked by the target user. The time period of each label is preset with a corresponding initial attenuation factor value. For example, the initial attenuation factor values at four time points of 0 to 3 days, 4 to 6 days, 7 to 15 days, and 16 or more days are 1, 0.8, 0.5, and 0.3, respectively. It is understood that different initial attenuation factor values can be set for different time periods, and in addition, the initial attenuation factor value corresponding to the time period in which the tag is currently located can be directly defined as eTime periodThe embodiment of the present invention is not particularly limited thereto.
And S55, respectively setting corresponding second initial values for the labels according to the sorting result.
Specifically, each tag in the tag combination may be divided into a plurality of levels according to a preset ratio according to the sorting result, and different second initial values may be set for tags in different levels. For example, the weight values of the labels corresponding to the business targets are sorted from large to small, the labels are divided into three levels according to a ratio of 3:4:3, and the labels in each level are set with different second initial values, for example, the first initial values set for the labels in the three levels are respectively 3,2 and 1. It will be appreciated that the earlier the level, the greater the initial value to which the tag in the level corresponds.
It should be noted that the first initial value and the second initial value corresponding to the same tag may be the same or different.
And S56, taking the second initial value corresponding to each label as an attenuation value, and performing attenuation calculation on the initial attenuation factor value corresponding to the time period in which each label is currently located to obtain the time attenuation factor value corresponding to the time period in which each label is currently located.
Specifically, the second initial value corresponding to each tag is correspondingly multiplied by the initial attenuation factor value corresponding to the time period in which each tag is currently located, so as to obtain the time attenuation factor value corresponding to the time period in which each tag is currently located.
And S57, calculating to obtain the time attenuation coefficient corresponding to each label according to the initial coefficient corresponding to each label and the time attenuation factor value corresponding to the time period in which each label is currently positioned.
Specifically, the initial coefficient corresponding to each tag is correspondingly multiplied by the time attenuation factor value corresponding to the time period in which each tag is currently located, so as to obtain the time attenuation coefficient corresponding to each tag. The embodiment of the present invention does not limit the specific calculation process.
In this embodiment, a corresponding preset attenuation value is set for a current time period of a service, a corresponding first initial value is set for each tag, an initial coefficient corresponding to each tag is obtained through calculation, a corresponding initial attenuation factor value is set for each current time period of each tag, a second initial value corresponding to each tag is used as an attenuation value, a time attenuation factor value corresponding to each current time period of each tag is obtained through calculation, and finally, a time attenuation coefficient corresponding to each tag is obtained through calculation by combining the initial coefficient corresponding to each tag and the time attenuation factor value corresponding to each current time period of each tag, so that the degree of influence of time on each tag can be more accurately reflected by the time attenuation factor corresponding to each tag.
It should be understood that, although the steps in the flowcharts of fig. 2 to 5 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-5 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performing the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
Fig. 6 is a block diagram of a structure of a user tag recommendation apparatus according to an embodiment of the present invention, and as shown in fig. 6, the apparatus may include:
a tag determining module 61, configured to determine a tag combination related to a business target from a tag set of a target user;
a first obtaining module 62, configured to obtain, according to a first algorithm, a weight value of each tag in the tag combination corresponding to the service target;
a second obtaining module 63, configured to obtain, according to a second algorithm, a weight value of each tag in the tag combination corresponding to the target user;
a third obtaining module 64, configured to obtain, according to a third algorithm, a time attenuation coefficient corresponding to each tag in the tag combination;
a weight calculating module 65, configured to calculate a final weight value of each tag based on a weight value of each tag corresponding to the target service, a weight value of each tag corresponding to the target user, and a time attenuation coefficient corresponding to each tag;
and the tag recommending module 66 is configured to sort the tags according to the respective final weight values, screen out effective tags, and recommend the effective tags to corresponding collectors in association with the service targets.
Further, when the hasten case associated with the target user is successfully promoted as the business target, the tag determining module 61 is specifically configured to:
acquiring a plurality of basic labels, a plurality of behavior labels and a plurality of interactive labels of a target user from a label set to form a label combination;
when the call number associated with the target user is used as the valid number as the service target, the tag determining module 61 is specifically configured to:
and acquiring a plurality of interactive labels of the target user from the label set to form a label combination.
Further, the first obtaining module 62 is specifically configured to:
determining a plurality of groups corresponding to each label, and performing WOE calculation and IV calculation on the plurality of groups of each label respectively to obtain an IV value of each label;
training a preset model by taking each label as an input variable and taking a business target as an output variable, and calculating to obtain a numerical value of the feature importance of each label;
and for each label, carrying out weighted summation on the IV value of the label and the numerical value of the characteristic importance of the label to obtain a weight value of the label corresponding to the business target.
Further, the second obtaining module 63 is specifically configured to:
for each tag, the following operations are performed:
calculating the TF value of the label according to the number of times of the label and the total number of times of all the labels in the label combination;
calculating the IDF value of the label according to the total number of all users and the number of the users containing the label;
and calculating the weight value of the label corresponding to the target user according to the TF value of the label and the IDF value of the label.
Further, the third obtaining module 64 is specifically configured to:
sorting the weighted values of all labels in the label combination corresponding to the business targets from large to small, and respectively setting corresponding first initial values for all labels according to sorting results;
determining a current time period of a service corresponding to a service target;
performing attenuation calculation on a first initial value corresponding to each label according to a preset attenuation value corresponding to a current service time period to obtain an initial coefficient corresponding to each label;
acquiring initial attenuation factor values corresponding to current time periods of all the tags, wherein the current time period of any tag is determined based on the number of days until the target user marks the tag for the first time;
respectively setting corresponding second initial values for the labels according to the sequencing result;
taking the second initial value corresponding to each label as an attenuation value, and performing attenuation calculation on the initial attenuation factor value corresponding to the time period in which each label is currently located to obtain the time attenuation factor value corresponding to the time period in which each label is currently located;
and calculating to obtain the time attenuation coefficient corresponding to each label according to the initial coefficient corresponding to each label and the time attenuation factor value corresponding to the time period in which each label is currently positioned.
Further, when the collection case associated with the target user is taken as a service, the third obtaining module 64 is specifically configured to:
determining the current time period of the business according to the overdue days of the collection urging case;
when the call number associated with the target user is used as a service, the third obtaining module 64 is specifically configured to:
and determining the current time period of the service according to the time days of the first connection of the call number.
Further, the tag recommendation module 66 is further configured to:
for each label, comparing the final weight value of the label with a preset threshold value;
and if the final weight of the label exceeds a preset threshold value, preferentially taking the label as a target label, and recommending the target label to a corresponding acquirer in a way of being associated with the business target.
The user tag recommendation device provided by the embodiment of the invention belongs to the same inventive concept as the user tag recommendation method provided by the embodiment of the invention, can execute the user tag recommendation method provided by the embodiment of the invention, and has the corresponding functional modules and beneficial effects of executing the user tag recommendation method. For technical details that are not described in detail in this embodiment, reference may be made to the user tag recommendation method provided in this embodiment of the present invention, and details are not described here again.
Fig. 7 is an internal structural diagram of a computer device according to an embodiment of the present invention. The computer device may be a server, and its internal structure diagram may be as shown in fig. 7. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a user tag recommendation method.
Those skilled in the art will appreciate that the architecture shown in fig. 7 is merely a block diagram of some of the structures associated with the inventive arrangements and is not intended to limit the computing devices to which the inventive arrangements may be applied, as a particular computing device may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
The embodiment of the invention provides computer equipment, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the computer program to realize the following steps:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to a business target according to a first algorithm;
acquiring a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating the final weight value of each label based on the weight value of each label corresponding to the target service, the weight value of each label corresponding to the target user and the time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
Furthermore, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the following steps:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to a business target according to a first algorithm;
acquiring a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating the final weight value of each label based on the weight value of each label corresponding to the target service, the weight value of each label corresponding to the target user and the time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, databases, or other media used in embodiments provided herein may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A user tag recommendation method, the method comprising:
determining a label combination related to a business target from a label set of a target user;
acquiring a weight value of each label in the label combination corresponding to the service target according to a first algorithm;
obtaining a weight value of each label in the label combination corresponding to the target user according to a second algorithm;
acquiring a time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
calculating a final weight value of each label based on a weight value of each label corresponding to the target service, a weight value of each label corresponding to the target user and a time attenuation coefficient corresponding to each label;
and sequencing the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
2. The method of claim 1,
when the collection case associated with the target user is successfully collected as the business target, the determining of the label combination related to the business target from the label set of the target user comprises:
acquiring a plurality of basic labels, a plurality of behavior labels and a plurality of interactive labels of the target user from the label set to form a label combination;
when the call number associated with the target user is taken as the effective number as the service target, the determining of the label combination related to the service target from the label set of the target user comprises the following steps:
and acquiring a plurality of interactive labels of the target user from the label set to form the label combination.
3. The method according to claim 1 or 2, wherein the obtaining the weight value of each label in the label combination corresponding to the service objective according to the first algorithm comprises:
determining a plurality of groups corresponding to the labels respectively, and performing WOE calculation and IV calculation on the groups of the labels respectively to obtain IV values of the labels;
training a preset model by taking each label as an input variable and taking the service target as an output variable, and calculating to obtain a numerical value of the feature importance of each label;
and for each label, carrying out weighted summation on the IV value of the label and the numerical value of the characteristic importance of the label to obtain a weight value of the label corresponding to the business target.
4. The method according to claim 1 or 2, wherein the obtaining the weight value of each tag in the tag combination corresponding to the target user according to the second algorithm comprises:
for each of the tags, performing the following operations:
calculating the TF value of the label according to the number of times of the label and the total number of times of all the labels in the label combination;
calculating the IDF value of the label according to the total number of all users and the number of the users containing the label;
and calculating the weight value of the label corresponding to the target user according to the TF value of the label and the IDF value of the label.
5. The method according to claim 1 or 2, wherein the obtaining the time attenuation coefficient corresponding to each tag in the tag combination according to a third algorithm comprises:
sorting the weighted values of the labels in the label combination corresponding to the business targets from large to small, and setting corresponding first initial values for the labels respectively according to the sorting result;
determining the current time period of the service corresponding to the service target;
performing attenuation calculation on a first initial value corresponding to each label according to a preset attenuation value corresponding to the current time period of the service to obtain an initial coefficient corresponding to each label;
acquiring initial attenuation factor values corresponding to time periods of the tags, wherein the time period of any one tag is determined based on the number of days of time for marking the tag for the first time by the target user;
respectively setting corresponding second initial values for the labels according to the sorting result;
taking the second initial value corresponding to each label as an attenuation value, and performing attenuation calculation on the initial attenuation factor value corresponding to the time period in which each label is currently located to obtain the time attenuation factor value corresponding to the time period in which each label is currently located;
and calculating to obtain the time attenuation coefficient corresponding to each label according to the initial coefficient corresponding to each label and the time attenuation factor value corresponding to the time period in which each label is currently positioned.
6. The method of claim 5,
when the collection urging case associated with the target user is taken as the service, the determining of the current time period of the service corresponding to the service target includes:
determining the current time period of the business according to the overdue days of the collection urging case;
when the call number associated with the target user is used as the service, the determining the current time period of the service corresponding to the service target includes:
and determining the current time period of the service according to the time days of the first connection of the call number.
7. The method of claim 1, further comprising:
for each label, comparing the final weight value of the label with a preset threshold value;
and if the final weight value of the label exceeds the preset threshold value, preferentially taking the label as the target label, and recommending the label to a corresponding acquirer in a way of being associated with the service target.
8. A user tag recommendation apparatus, the apparatus comprising:
the label determining module is used for determining a label combination related to the business target from the label set of the target user;
the first obtaining module is used for obtaining the weight value of each label in the label combination corresponding to the service target according to a first algorithm;
the second obtaining module is used for obtaining the weight value of each label in the label combination corresponding to the target user according to a second algorithm;
the third obtaining module is used for obtaining the time attenuation coefficient corresponding to each label in the label combination according to a third algorithm;
a weight calculation module, configured to calculate a final weight value of each tag based on a weight value of each tag corresponding to the target service, a weight value of each tag corresponding to the target user, and a time attenuation coefficient corresponding to each tag;
and the label recommending module is used for sorting the labels according to the respective final weight values, screening out effective labels, and recommending the effective labels and the service targets to corresponding receivers in a correlated manner.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the user tag recommendation method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the user tag recommendation method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911167962.5A CN111061948B (en) | 2019-11-25 | 2019-11-25 | User tag recommendation method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911167962.5A CN111061948B (en) | 2019-11-25 | 2019-11-25 | User tag recommendation method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111061948A true CN111061948A (en) | 2020-04-24 |
| CN111061948B CN111061948B (en) | 2023-05-23 |
Family
ID=70298652
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911167962.5A Active CN111061948B (en) | 2019-11-25 | 2019-11-25 | User tag recommendation method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111061948B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112732934A (en) * | 2021-01-11 | 2021-04-30 | 国网山东省电力公司电力科学研究院 | Power grid equipment word segmentation dictionary and fault case library construction method |
| CN112734463A (en) * | 2020-12-30 | 2021-04-30 | 咪咕音乐有限公司 | Service information sending method and device, electronic equipment and storage medium |
| CN112991049A (en) * | 2021-04-13 | 2021-06-18 | 重庆度小满优扬科技有限公司 | Loan information processing method and electronic device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017133456A1 (en) * | 2016-02-01 | 2017-08-10 | 腾讯科技（深圳）有限公司 | Method and device for determining risk evaluation parameter |
| CN109509086A (en) * | 2018-11-28 | 2019-03-22 | 上海点融信息科技有限责任公司 | The method, apparatus and storage medium of processing collection business based on artificial intelligence |
| CN109815489A (en) * | 2019-01-02 | 2019-05-28 | 深圳壹账通智能科技有限公司 | Collection information generating method, device, computer equipment and storage medium |
| CN110033165A (en) * | 2019-03-06 | 2019-07-19 | 平安科技（深圳）有限公司 | The recommended method of overdue loaning bill collection mode, device, medium, electronic equipment |
-
2019
- 2019-11-25 CN CN201911167962.5A patent/CN111061948B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017133456A1 (en) * | 2016-02-01 | 2017-08-10 | 腾讯科技（深圳）有限公司 | Method and device for determining risk evaluation parameter |
| CN109509086A (en) * | 2018-11-28 | 2019-03-22 | 上海点融信息科技有限责任公司 | The method, apparatus and storage medium of processing collection business based on artificial intelligence |
| CN109815489A (en) * | 2019-01-02 | 2019-05-28 | 深圳壹账通智能科技有限公司 | Collection information generating method, device, computer equipment and storage medium |
| CN110033165A (en) * | 2019-03-06 | 2019-07-19 | 平安科技（深圳）有限公司 | The recommended method of overdue loaning bill collection mode, device, medium, electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 陶红燕;胡蔚霞;沈艳;: "运用多种催收策略有效控制信用卡风险" * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112734463A (en) * | 2020-12-30 | 2021-04-30 | 咪咕音乐有限公司 | Service information sending method and device, electronic equipment and storage medium |
| CN112732934A (en) * | 2021-01-11 | 2021-04-30 | 国网山东省电力公司电力科学研究院 | Power grid equipment word segmentation dictionary and fault case library construction method |
| CN112991049A (en) * | 2021-04-13 | 2021-06-18 | 重庆度小满优扬科技有限公司 | Loan information processing method and electronic device |
| CN112991049B (en) * | 2021-04-13 | 2023-05-30 | 重庆度小满优扬科技有限公司 | Loan information processing method and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111061948B (en) | 2023-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110837931B (en) | Customer churn prediction method, device and storage medium | |
| CN109165840B (en) | Risk prediction processing method, risk prediction processing device, computer equipment and medium | |
| CN110070391B (en) | Data processing method and device, computer readable medium and electronic equipment | |
| CN108876600A (en) | Warning information method for pushing, device, computer equipment and medium | |
| CN112633962B (en) | Service recommendation method and device, computer equipment and storage medium | |
| CN111061948B (en) | User tag recommendation method and device, computer equipment and storage medium | |
| CN113627566A (en) | Early warning method and device for phishing and computer equipment | |
| CN111738819A (en) | Method, device and equipment for screening characterization data | |
| KR20200075120A (en) | Business default prediction system and operation method thereof | |
| CN110634060A (en) | User credit risk assessment method, system, device and storage medium | |
| CN110728301A (en) | Credit scoring method, device, terminal and storage medium for individual user | |
| CN111695084A (en) | Model generation method, credit score generation method, device, equipment and storage medium | |
| CN112966189A (en) | Fund product recommendation system | |
| CN111967807A (en) | Method and device for generating risk event judgment rule executed by computer | |
| CN112836750A (en) | System resource allocation method, device and equipment | |
| CN112232950A (en) | Loan risk assessment method and device, equipment and computer-readable storage medium | |
| CN115034886A (en) | Default risk prediction method and device | |
| CN109146667B (en) | Method for constructing external interface comprehensive application model based on quantitative statistics | |
| CN113887214A (en) | Artificial intelligence based wish presumption method and related equipment thereof | |
| WO2021098652A1 (en) | Data processing method and device | |
| CN114997879B (en) | Payment routing method, device, equipment and storage medium | |
| CN117196630A (en) | Transaction risk prediction method, device, terminal equipment and storage medium | |
| CN112084408B (en) | List data screening method, device, computer equipment and storage medium | |
| TWI792101B (en) | Data Quantification Method Based on Confirmed Value and Predicted Value | |
| CN114925275A (en) | Product recommendation method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231211 Address after: Room 811, 8th Floor, No. 20, Lane 999, Dangui Road, China (Shanghai) Pilot Free Trade Zone, Pudong New Area, Shanghai, March 2012 Patentee after: Shanghai Platinum Zinc Information Technology Co.,Ltd. Address before: No. 8, 1st Floor, Building 1, No. 39 Renhe Street, Chengdu High tech Zone, Chengdu, Sichuan Province, 610000 Patentee before: Chengdu platinum tin Financial Information Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |